This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. Primary support for the subproject and the subproject's principal investigator may have been provided by other sources, including other NIH sources. The Total Cost listed for the subproject likely represents the estimated amount of Center infrastructure utilized by the subproject, not direct funding provided by the NCRR grant to the subproject or subproject staff. (A) OBJECTIVES Atrial fibrillation (AF) affects 2.2 million individuals in the United States, and is a major cause of stroke, heart failure and mortality (8). Maintaining sinus rhythm reduces symptoms and may prolong survival, yet remains difficult (1). Recent advances in ablation now make it possible to cure many patients with paroxysmal AF, whose episodes are self-limiting, by isolating triggers in the pulmonary veins (12, 14). Unfortunately, ablation is complex, and less successful in the large population with persistent AF, whose episodes require drugs or cardioversion to terminate (10). In this group, AF recurs post-ablation in >50 %, requiring multiple ablations with attendant morbidity and mortality (5). Unfortunately, therapy is limited by a poor understanding of how, and under what conditions, AF occurs in humans. It is increasingly appreciated that structural heterogeneities or dynamic tissue properties may initiate fibrillation (33). Animal, in vitro and computational studies have shown that every-other-beat oscillations (alternans) in action potential duration (APD), may initiate ventricular fibrillation (33). Mechanistically, tissue heterogeneities such as scar, or dynamics such as steep restitution (i.e. "rate-response") of APD, conduction velocity (CV) that slows for a broad range of rates (36), may cause APD alternans. By exaggerating repolarization dispersion, particularly if discordant (33), APD alternans may be a direct mechanism for AF. Although APD alternans has yet to be linked with AF in animals or man, we have exciting preliminary data in humans showing APD alternans leading directly to reentrant AF. As a collaborative project of the NBCR this research will promote extensions to the development of Continuity and its anatomic and electrical models and patient-specific modeling algorithms that will permit greater integration with models of impulse conduction in the atria. Our central hypothesis is that Atrial Fibrillation in humans initiates from Alternans of Action Potential Duration (APD), that reflects steep restitution of atrial APD and broad restitution of regional conduction velocity, and explains AF near the pulmonary veins (PV) in paroxysmal AF but not persistent AF. This study marries sophisticated data collection in patients at AF ablation with unique state-of-the-art patient-specific computational modeling to address 3 Specific Aims. 1. To determine whether alternans of atrial action potential duration (APD), resulting from steep APD restitution or broad conduction velocity (CV) restitution, precedes the onset of Atrial Fibrillation. We will record multi-site monophasic action potentials (MAP) and CV from 64-128 bi-atrial basket poles at electrophysiologic study, with and without pharmacologic modulation, in atrial reconstructions guided by computed tomography in paroxysmal and persistent AF patients. 2. To determine whether the first beats of AF follow conduction block and reentry. We will use patient-specific structure-function data, from basket maps referenced to digital atrial anatomy, isochronal analysis and phase mapping. We will also determine if these sites lie near PVs in patients with paroxysmal AF but not persistent AF. 3. To determine whether AF is caused by atrial discordant APD alternans, by developing patient-specific computational models derived from clinically observed electrophysiology. We will develop finite-volume models that incorporate observed CV and APD restitution, atrial shape and structural heterogeneities for each patient, to compare modeled to actual AF in each patient.